MdAliAhnaf/Skin_Type_Classification-Recommendation ? reverse-engineered prompt

Build me a small skin photo analyzer that can look at an image and tell whether the skin looks acne prone, dry, or oily.

I want it to train and test a few image classification models with transfer learning, then compare which one works best and show the results clearly with charts and metrics like accuracy, precision, recall, and F1 score. Use the skin image dataset already in the project, and make sure the training process saves the model and the training history.

After it predicts the skin type, I want personalized skincare recommendations generated from that result, and if possible let it use OpenAI for the advice. It should also be easy to run from notebooks, since I plan to test everything there. If you need to look up current docs for anything, go ahead and do that.